Estimation and Analysis for Geographic and Orographic Influences on Precipitation Distribution in China

Shu Shou-Juan,Wang Yuan,Xiong An-Yuan
DOI: https://doi.org/10.1002/cjg2.1168
2007-01-01
Chinese Journal of Geophysics
Abstract:Different from the framework of multiple linear regression based on least square method, this paper tries to apply a novel second-generation regression method based on partial least-squares to precipitation estimation in China for the first time. The 45-yr precipitation data from 394 meteorological stations in study area are used. Several simple formulae used to estimate the annual mean and seasonal precipitation have been obtained, and the characteristics of the geographic or topographic effects have been presented. The impact factors include longitude, latitude, height, slope, sloping direction and close limit. The results show that the fraction of the variation of response explained by the model is above 70%, and the average correlation coefficients are nearly all above 0.84. The results are satisfied through the test of cross-validation. Through it is not appropriate to set up multiple linear regression models, the estimated precipitation based on partial least-squares regression correctly replicates real spatial distribution of precipitation qualitatively and quantitatively.
What problem does this paper attempt to address?